Racial differences in health care are documented across the health care continuum and persist in aging and end-of-life (EOL) care. African Americans (AA) and other underrepresented minorities often choose more aggressive therapies at the end of life and are less likely to utilize hospice care in the terminal stages of their illness. Potential reasons for these disparities include: lack of knowledge of and misperceptions about palliative and hospice care, spiritual beliefs, and mistrust in the health care system, among others. Despite the literature on disparities in end-of-life (EOL) care and reasons for underuse and the presence of national EOL care guidelines, attempts to address this problem have been limited and often not rigorously evaluated. The majority of interventions to promote EOL care were done in majority populations and focused predominantly on trying to change physician awareness of patient's pain, symptoms, and values or to change physician communication behavior. While these early studies made tremendous contributions to the study of EOL care and the needs of the terminally ill, the interventions associated with these studies did not reach their desired effectiveness. The investigators propose a different strategy that would focus specifically on previously identified barriers to utilization of advance directives, palliative care, and hospice care among African Americans - including physicians' difficulty and discomfort with prognostication, AA patients' knowledge, attitudes and beliefs towards hospice and palliative care, conflict between patients' spiritual beliefs and the general hospice and palliative medicine philosophy of care, and medical mistrust. The goal of this project is to improve methods of prognostication for physicians and increase awareness of EOL care options for AAs. To overcome the dual challenges of physicians' reluctance to discuss EOL care and patients' discomfort in engaging in such conversations, the investigators will use the electronic medical record (EMR) to automatically identify AA patients with life-limiting illness who are eligible for counseling about EOL care options. To change knowledge and attitudes toward EOL care options among AA patients, the investigators will design a culturally sensitive intervention that will combine multimedia materials and a culturally concordant lay health advisor who will deliver tailored education and counseling.

Aim 1:

A. We will conduct semi-structured interviews with patient/caregiver pairs and focus groups with providers that will test the communication strategies of available audiovisual materials and materials that we will develop for the intervention. The audiovisual segments will be taken from an available EOL care educational digital versatile disc (DVD). We aim to interview 12 patients and their caregivers, or more until thematic saturation is reached. We will obtain informed consent, and all interviews will be audiotaped and transcribed. Interviews will be conducted separately, and will last 30 to 45 minutes.

B. Two focus groups will be conducted with palliative care providers to identify communication strategies they use with AAs. One will be conducted with providers from Parkland Hospital. The other will be conducted with providers from University Hospitals, the Dallas VA, and Baylor University Medical Center in Dallas. Informed consent will be obtained. All sessions will be audiotaped and transcribed. The focus groups will last 45 to 60 minutes.

C. We will create additional DVD segments that will address previously identified barriers to EOL care for AAs, including: 1) spiritual/religious conflict, and 2) medical mistrust. We will obtain feedback on the newly developed segments from 10 new patient/caregiver dyads (semi-structured interviews) and 6 to 8 AA religious leaders (focus group) recruited from local churches. Informed consent will be obtained prior to conducting the interviews and focus groups. All will last 30 to 60 minutes.

D. A lay health advisor (LHA) will be recruited from the community to provide counsel about EOL care to patients who will receive the intervention. The LHA and the PI of the project will undergo training in how to provide culturally sensitive EOL care for AAs via the APPEAL (A Progressive Palliative Care Educational Curriculum for the Care of AAs at Life's End) Curriculum created at Duke University. After training, the LHA will participate in in-service work with the Parkland Palliative Care Team and the PI.

Aim 2:

We will use an e-EOL algorithm to identify AA patients hospitalized at Parkland who have advanced breast, lung, and colorectal cancer to identify potentially eligible candidates for the intervention utilizing EMR data from Parkland Hospital (See Aim 1 eligibility criteria). Once eligibility is confirmed the LHA will introduce the study to the patient and obtain informed consent. Each patient will be asked to identify a primary caregiver that will be able to participate in the intervention. The LHA will contact the patients' primary caregivers to confirm participation in the study and arrange a time to meet with both the patient and caregiver to conduct the intervention. We anticipate that 24 patient-caregiver pairs will receive the intervention (8 for each type of cancer).

The LHA will meet with eligible patients and caregivers and assist them in watching the developed DVD segments. Afterward, the LHA will answer questions and provide additional information. They will tailor the discussion to the patient's values, preferences, concerns, and clinical circumstances.

The primary process outcome tested will be the feasibility and acceptability of the intervention. Feasibility success will be measured by the number and rates of patients/caregivers who complete the intervention and follow-up interviews. The primary decision-making outcome is change in intent to discuss EOL care options based on the Transtheoretical Stages of Change Model (i.e., pre-contemplation, contemplation, preparation, and action). Secondary outcomes measured will include: knowledge of prognosis and EOL care options, decisional conflict, quality of life, and health care utilization. Other patient and treatment variables will also be collected, per the study protocol.

Inclusion Criteria:

Aim 1 patients must:

1. receive their care at Parkland and be diagnosed with advanced cancer (breast, lung, or colon);
2. self-identify as AA;
3. be proficient in English;
4. be competent to give informed consent; and
5. have no evidence of cognitive impairment (Mini-Cog score of ≥3 or 1-2 with normal clock draw).

Aim 2 Patients must:

1. be hospitalized at Parkland
2. be diagnosed with advanced cancer (breast, lung, or colon)
3. self-identify as AA;
4. be proficient in English;
5. be competent to give informed consent;
6. have no evidence of cognitive impairment (Mini-Cog score of ≥3 or 1-2 with normal clock draw); and
7. have never received palliative or hospice care.

All Caregivers (Aim 1 and 2) must be:

1. identified by the selected patients as their primary caregiver;
2. be 21 years of age or older;
3. proficient in English; and
4. competent to give informed consent.

For the expert provider focus group, participants must be a health care provider (physician, nurse practitioner, chaplain, social worker, nurse) who works within hospice and palliative medicine.

(Note: patients who enter palliative care or hospice during follow-up interviews will be allowed to remain in the study)

Exclusion Criteria:

For patients:

1. identify with a race other than African American or
2. have a diagnosis other than advanced breast, lung, or colorectal cancer.

One patient initially assigned to the control group was deemed ineligible and was removed from the study, leaving 12 patients assigned to the control group for analysis.

1 year

The study procedures do not involve any medical procedures or interventions. Procedures are limited to patients answering survey questions, and patients randomized to the intervention arm will view an educational video. Because eligible patients must have a diagnosis of advanced cancer, it's possible some patients my die prior to completing study surveys ("all-cause mortality"). In addition, patients may experience psychological distress answering survey questions ("other").

The research staff conducting assessments for all participants was masked to patients receiving the intervention as well as those assigned to the control group.